.

I will describe two projects going on in my lab: 1) CACTUS: integrating clonal architecture with genomic clustering and transcriptome profiling of single tumor cells and 2) Interpretable deep recommender system model for prediction of kinase inhibitor efficacy across cancer cell lines. Below I provide a short summary for both of them.1) CACTUS: integrating clonal architecture with genomic clustering and transcriptome profiling of single tumor cells  Drawing genotype-to-phenotype maps in tumors is of paramount importance for understanding tumor heterogeneity. Assignment of single cells to their tumor clones of origin can be approached by matching the genotypes of the clones to the mutations found in RNA sequencing of the cells. CACTUS is a probabilistic model that leverages the information from an independent genomic clustering of cells and exploits the scarce single cell RNA sequencing data to map single cells to given imperfect genotypes of tumor clones. We apply CACTUS to two follicular lymphoma patient samples, integrating three…

To understand the sequence relationships between many genomes, we build a graphical data structure that represents their mutual alignment. This pangenome graph can serve as a powerful tool for research into genome variation, selection, and evolution. Generating and refining pangenome graphs is an area of active research that has necessitated the development of novel alignment and graph induction methods. We discuss these issues in detail and illustrate how they are addressed in the pangenome graph builder tool (pggb). We close with an overview of the application of our approach to the 90 haplotypes in the year 1 assemblies of the Human Pangenome Project, where our method provides an unprecedented view of variation in the full human pangenome.

Cancer is a genetic disease, where cell division, mutation and selection produce a heterogeneous tumor composed of multiple subpopulations of cells with different sets of mutations. During later stages of cancer progression, cancerous cells from the primary tumor migrate and seed metastases at distant anatomical sites. The cell division and mutation history of an individual tumor can be represented by a phylogenetic tree, which helps guide patient-specific treatments. In this talk, I will introduce combinatorial algorithms for reconstructing tumor phylogenies from bulk DNA sequencing data, where the measurements are a mixture of thousands of cells. These algorithms are based on a combinatorial characterization of phylogenetic trees as a restricted class of spanning trees in a graph, a characterization that also demonstrates the computational complexity of the problem. In addition, I will introduce a novel framework for analyzing the history of cellular migrations between anatomical sites in metastatic cancers. Finally, I will discuss…

Genetically identical individuals growing in the same environment often show substantial phenotypic variation. The reasons for this phenotypic variation are usually unknown. We start from genome wide expression profiling on genetically identical C. elegans roundworms to characterize molecular differences, understand their phenotypic impact and infer the mechanisms that generate such variation. I will show how using this strategy we previously found maternal age to be an important determinant of phenotypic variation with changes in maternal provisioning of yolk to the embryo explain some of the observed phenotypic differences. Then I will show we have now found the ‘smell’ of the worms (i.e pheromones) – which is perceived by neurons – also induces important phenotypic variation generating an inherited memory that alters germline development and minimum generation time in the progeny of exposed worms. 

Dimensional reduction approaches are crucial to extract relevant information from high-dimensional datasets, such as clinical or omics data. We have developed a toolbox based onnon-negative matrix factorization, in order to extract relevant biological and clinicalsignatures from high-dimensional datasets and combine various modalities into molecularsignatures. I will present some recent applications of these approaches, fromneuroblastoma epigenomics to the study of comorbidities between mental and somaticdisorders and the application to transfer learning from single-cell to bulk datasets.

The problem of controlling a dynamic network has a long history in control theory, with roots in a diversity of mathematical methods from complex analysis to topology, graph theory and computational complexity. The basic problem setup is that of a dynamical system represented as a directed graph, with nodes influencing each other’s dynamics. Control is sought over a given set of targets, in the sense of being able to change their configuration through external interventions on some well-chosen input nodes in the network, taking advantage of the network topology. We are interested in finding a minimal set of input nodes in the network such that the behavior of the target nodes may be changed arbitrarily through a well-chosen sequence of signals to the input nodes, cascaded throughout the network through its wiring. We focus on formalizations of this network controllability problem that maximize its applicability in biomedicine, including a specific set of…